Design Of Campus Card Query System And Analysis Of Implicit User Behavior

With the rapid development of information technology to promote the construction of Digital Campus, campus card system has been widely used in the major colleges and universities across the country covering all aspects of life and learning. The system generates a lot of data every day and stores it in the database, 10 million or even hundreds of billion data has been accumulated after many years of application. These data hide the behavior characteristics of users, how to use massive amounts of data to provide decision-making support for the daily management of the school is an urgent problem to be solved. In order to comply with the development of information technology in colleges and universities, this paper uses data mining technology to analyze the student card data. This paper completed the following work:(1) Design and build a campus card query system based on J2EE platform. In order to realize the development of visualization system later, using Struts2+Wevservice+J-UI framework to develop campus card query system under the premise of the school providing a Webservice interface, the majority of teachers and students can inquire their own consumption details and subsidies and other information expediently to achieve self-service and informatization business through the system.(2) Campus card basic information visualization. Choosing the information of the canteen supermarket consumption and the library access doing the statistics of the school students breakfast time, the number and amount of canteen and supermarket credit card, the number of credit card library, analyzing the characteristics of different gender groups in each grade to intuitively learn about the data and make card data mining analysis later convenient. Doing the statistics of canteen card consumption frequency and the number of diners in different periods of time, analyze the characteristics of peak and volume rate for canteen dinner.(3) User portrait. This part uses the data mining technology to carry on user portrait analysis of the students campus consumption and learning activities data.Firstly, extracting key features through the data pretreatment, using K-means clustering algorithm to cluster the undergraduate data set after standardized treatment,illustrating portrait after analyzing the characteristics of the consumption habits of users and the crowd. Finally, classify undergraduate data set through the generated decision tree model to evaluate the accuracy of the population characteristic parameters. The practical results show the user classification model designed in this paper can effectively distinguish the users with different behavior characteristics to provide the basis for the management of College students.(4) Canteen turnover forecast. In this part, use the data of the campus card to analyze the business volume law of the canteen and study the application of the canteen consumption data in the time latitude to provide solution for the promotion of the digital campus development and the establishment of the forecasting and early warning system. Constructing time series forecasting model based on the daily turnover data of the canteen, comparing the actual value and the predicted value of the canteen turnover to verify the rationality of the model.